ANN based Load Prediction in Residential Buildings

Abstract: Load prediction of residential buildings is a major area in which research has been going on.The primary aim is to improve the efficiency of demand prediction models. The conventional systems using fixed set point based products like SCADA is not seen to be giving the desired results. In such models the large gap between actual and predicted values of energy consumption causes huge error and it makes such systems unsuitable for use. In such a situation the use of thermal energy model with ANN(artificial neural network)is found to be giving better results with less gap in energy demand and prediction. This project is aimed to study the energy consumption pattern of residential buildings in kerala. Few physical parameters are considered in developing the the building model

Keywords: Load Prediction, SCADA, ANN, TOD, DISCOM.